Natural language analysis to support biomimetic design

Biomimetic design relies on relevant biological phenomena to serve as a basis for concepts in the engineering domain. Many instances of biomimetic design have resulted from personal observations of biological phenomena. However, a non-biologist's biology knowledge may be limited. To overcome this limitation, we perform functional keyword searches on existing natural-language knowledge sources. However, differences in domain lexicons present challenges to retrieving relevant information. A meaningful keyword to an engineer may not result in relevant matches within the biology domain.; This thesis presents studies that develop a method to systematically bridge these disparate domains. The results suggest that language analysis can be used to facilitate the biomimetic design process by providing the designer with biologically meaningful keywords that may not be intuitively obvious. The ability to find these keywords, and thus related biological phenomena, encourages creativity within the conceptual design process.